defuddle

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). This skill (SKILL.md) explicitly fetches and cleans arbitrary public web pages (e.g., the "defuddle https://example.com/article" usage and the "Integration with /wiki-ingest" note that the ingest skill will call it when a URL is passed), meaning the agent will ingest untrusted third‑party content from open websites which could contain instructions that influence subsequent actions.